Holistic Modeling and Analysis of Multistage Manufacturing Processes with Sparse Effective Inputs and Mixed Profile Outputs Andi Wang Arizona State University Website: https://web.asu.edu/andi-wang Email: [email protected]August 6, 2021 1 Wang, A., & Shi, J. (2021). Holistic modeling and analysis of multistage manufacturing processes with sparse effective inputs and mixed profile outputs. IISE Transactions, 53(5), 582-596.
24
Embed
Holistic Modeling and Analysis of Multistage Manufacturing ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Holistic Modeling and Analysis of Multistage Manufacturing Processes with Sparse Effective
1Wang, A., & Shi, J. (2021). Holistic modeling and analysis of multistage manufacturing processes with sparse effective inputs and mixed profile outputs. IISE Transactions, 53(5), 582-596.
Multistage: โข Error propagationData-rich: โข Mixed profilesโข Redundant information in ๐ฎ"โs
DepositionWaferSubstrate Lithography Etching Polish Polish Deposition Final Product
Etching profileOverlay fieldFilm Thickness Film Thickness
Process and Data Characteristics
3
Cascading effects
โข Potential causes may affect current and later stages
Sparsity of actual root causes
โข Not all potential causes affect the quality simultaneously
Smoothness of quality measurements
โข Profiles of smooth curves and images
Few underlying quality issues
โข Actual root causes link to few latent quality issues and
thus cause limited variation patterns of profiles
Common Existing Approaches
4
Separate models [e.g., 5,6]Risk missing key features.
Two-step approach [e.g., 7]May not reach consensus on the actual root causes.
Stream-of-variation [1,2]โState vectorsโ are not well defined.
[1] Shi, J. (2006). Stream of variation modeling and analysis for multistage manufacturing processes. CRC press.[2] Shi, J., & Zhou, S. (2009). Quality control and improvement for multistage systems: A survey. IISE Transactions, 41(9), 744-753.[3] Ju, F., Li, J., Xiao, G., Huang, N., & Biller, S. (2013). A quality flow model in battery manufacturing systems for electric vehicles. IEEE Transactions on Automation Science and Engineering, 11(1), 230-244.[4] Ju, F., Li, J., Xiao, G., & Arinez, J. (2013). Quality flow model in automotive paint shops. International Journal of Production Research, 51(21), 6470-6483.[5] Lin, T. H., Hung, M. H., Lin, R. C., & Cheng, F. T. (2006, May). A virtual metrology scheme for predicting CVD thickness in semiconductor manufacturing. In Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006. (pp. 1054-1059). IEEE.[6] Ma, X., Zhao, Q., Zhang, H., Wang, Z., & Arce, G. R. (2018). Model-driven convolution neural network for inverse lithography. Optics express, 26(25), 32565-32584.[7] Moyne, J., & Iskandar, J. (2017). Big data analytics for smart manufacturing: Case studies in semiconductor manufacturing. Processes, 5(3), 39.
Quality Flow Model [3,4]Not suitable for data-rich MMPs.
5
Research Objective
Objective:To develop a holistic framework for the MMP in data rich environment
Unique contributions:
โข Analyze heterogeneous quality measures and potential root causes
simultaneously under a unified framework.
โข Develop a solution procedure based on distributed optimization
Which are the actual root causes that affect the process quality?
How these root causes affect the quality measurements?
How to associate the quality measures with several key variation patterns?